workflows/workflows/agent-environment-setup/platforms/antigravity/skills/stitch-design-orchestrator/SKILL.md
Orchestrate safe, design-first Google Stitch generation and editing by sequencing frontend-design, canonical design-system sync, prompt enhancement, Stitch MCP tool selection, and implementation handoff.
npx skillsauth add cubetiq/cubis-foundry stitch-design-orchestratorInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Coordinate Google Stitch work as a workflow-first, skill-driven sequence. This skill ensures UI generation or editing does not jump straight into remote Stitch calls without first establishing design direction, compact prompt structure, canonical design-system context, and an implementation plan for the repo's real stack.
frontend-design - Always establish the style contract, direction name, and thesis set before forming a Stitch prompt.design to choose the owning execution surface and decide whether the design language must be refreshed before Stitch runs.docs/foundation/DESIGN.md is missing, stale, or the request spans multiple screens, run design-system first and mirror the resolved state to the Stitch-facing design context only after the canonical file is ready.stitch-prompt-enhancement before any Stitch tool call - Convert rough user intent into a compact structured brief that names the platform, layout, components, visual mood, and change scope.stitch_get_status, mcp_gateway_status, and stitch_list_enabled_tools before choosing a tool flow. If Stitch is unavailable, stop treating it as authoritative input.list_projects before create_project and reuse an existing project when it already represents the current feature or product area. Call list_screens before planning follow-up work so edits stay attached to the current screen set instead of spawning unnecessary new artifacts.generate_screen_from_text for a net-new screen, edit_screens for targeted revisions, generate_variants for controlled alternatives, and create_design_system or apply_design_system only when the design system itself is the current task. Default to GEMINI_3_1_PRO for complex new screens, multi-screen work, and design-system-heavy tasks. Use GEMINI_3_FLASH only when the user explicitly wants a speed-first draft or the task is a lightweight edit.edit_screens over full regeneration once a screen exists, and stop after two automatic retries with backoff.generate_screen_from_text and generate_variants, a timeout can still leave a finished screen in Stitch. Check list_screens before assuming the generation failed, then continue from the recovered screen instead of creating a new project.get_screen after the final generation or edit pass so the downstream implementation step receives the actual latest artifact, not a guessed description.stitch-implementation-handoff - Once the design output is settled, hand off the artifact so the repo implementation reuses local components, tokens, and architecture.Deliver:
| File | Load when |
| --- | --- |
| references/tool-selection.md | Need to choose the right Stitch tool path for new screens, edits, variants, or design-system work. |
| references/anti-abuse.md | Need retry, backoff, or prompt-budget rules before using the remote Stitch service. |
| ../design/references/visual-direction.md | Need to define the interface mood and composition before writing a Stitch prompt. |
| ../design/references/design-tokens.md | Need semantic token language before sending UI instructions to Stitch. |
| ../design/references/execution-contract.md | Need the modern split between public intake, critique, systemization, and execution surfaces. |
| File | Use when |
| --- | --- |
| examples/01-stitch-ui-route.md | Routing a new Stitch UI request through the full design-first sequence. |
tools
Use when investigating latest vendor behavior, comparing tools or platforms, verifying claims beyond the repo, or gathering external evidence before implementation.
documentation
Use when designing database schemas, normalization strategies, indexing plans, query optimization, and migration workflows for relational, document, or hybrid data stores.
development
Use when writing, reviewing, or refactoring modern C#/.NET code, including minimal APIs, records, async streams, pattern matching, DI lifetimes, and memory-efficient performance tuning.
development
Use when conducting code reviews, building review checklists, calibrating review depth, providing structured feedback, or establishing team review practices. Covers review methodology, feedback patterns, automated checks, and batch review strategies.